Informatica Inc. (INFA) Earnings Call Transcript & Summary

So I like to be able to welcome Informatica back to the Goldman Sachs Communacopia and Technology Conference. Amit Walia, CEO; and Michael, Mike McLaughlin, you go by Mike?

Mike easily.

Mike, CFO, joined by my able co-moderator, Matt Martino, who will ask you the tough zinger questions.

All right.

We have good questions. Welcome to the conference.

Thank you for hosting. Appreciate it. Absolutely.

This is something that I think you can knock it out of the ballpark. How is -- and I think you know that I actually met Informatica's founder in 1995. So that's the old Informatica.

It's very old, for sure.

Very old Informatica. All 2, yes. So the new Informatica. Tell us what is the Informatica. So what are the things that are driving your growth, and what do you make of the new profile of the company versus the old?

Yes, absolutely. But thanks for hosting us, Kash. Look, I think the best way to explain that is, and if I can't compress -- and I can speak fast, so let me know if I'm too slowly. Look, 2015, we went private. And circa before 2015, this was a classic on-prem license software company, primarily driven by a single product power center, driving more ETL use cases for data warehouses. It's $1 billion revenue that time, half license, half maintenance. And then we had pieces at that point, and I run product to that point that look, the -- if you take the digital transformation thesis that play out over the course of time, 3 things will come true. One is it's not just growth of data. Many, many more use cases for data, which means a bigger TAM and many more products than we had before. Number two is the world will become cloud first. But until then, it will be hybrid, [ ground ] cloud. And then at that point, it will become multi-cloud. So it will remain hybrid for that -- in that transformation. And the third was that I had spent a fairly good time at security before, and it was a very fragmented industry. And same goes for data management. And the thesis we had is at that point in time, if you can be a platform provider, you will have a much more economic gains in the industry than ever before. So we set out to do that. And for that, we took the old product, kept it on the side, did not touch it, and we started building out the new product, new use cases, first, some on-prem and then cloud. Fast forward, as we come here today. Since 2015 -- 2016 to 2024, we are in 2024, 8 years. And if you look at our new product business, that's about $1.2 billion of ARR. This is 30th year of Informatica. So 22 years it took to build a $1 billion business with license software. And in 8 years, we built 1.2 billion new cloud -- new ARR business, of which 800 and change will be cloud this year. And that's the innovation. Now of course, when you add it up mathematically, of course, we are all math people in the room. So weighted average means that if this is growing very fast, something we don't sell anymore is declining, you get to single-digit returns. And in this new world, we are -- what we are very clear is we spent $1 billion in R&D to build, number one, the best products. Not a single Magic Quadrant or a [ Forrester ] has this number, too. Second is the platform story. We're the only platform provider in the industry of data management. And we have our AI baked into that, which by the way, [ IGA'd ] in 2018. Now with Gen AI, it's already live. And the third is, we believe that we will remain the Switzerland of data for any multi-vendor, multi-cloud, any kind of use case. And that's what we did. And I'll cap it off by saying that if you look at our cloud platform, and I know we'll touch on that later, is that that's running at industrial grade. We just finished our first half of the year, beat, raised our guide. And basically, the cloud platform, I will talk about ARR later. 97 trillion transactions a month is the scale of that platform that is growing at 60-plus percent every year.

I don't think we'd get to 97 trillion. That's a big number. What happened? How are the other 3 trillion -- what went missing?

We can round it 100. If you were -- I'll joke it. We are the only public company in data management . If I were a private company, I can say anything.

That's true. That's true. So talk about your bold audacious goals for the company 5 years out.

Well, look, I think we have set out to be the goal. We're going to be the only platform company in the place. There is just no one around us. Everybody is a bite-sized company. And our goal is to be the backbone of data-led blank transformation for the many years to come. It was data-led digital transformation. It's morphing into data-led AI transformation. And I believe at the scale, scope, breadth that we can do, nobody else can do. And philosophically, by the way, -- and our business has done very well, and I think the AI journey, and we'll talk about that later as well, is one of the best thing that's happening for data management because the fragmentation within an enterprise, people constantly think, oh, enterprise is only one database. Only that's not true. Enterprises are supremely more fragmented than before, and we help bring structure to that chaos, and we look at that. And I think last I'll say in this AI world, everybody is ready for AI except data, and we see that every day. And I think that's a massive tailwind for us.

That's great. So if you were to talk to the financial cloud 5 years out, I know you don't need to give guidance, but what are the other non-technological ways in which you would define success for Informatica? Like to be top something, top 4, top 5 in mid-cap? Is there anything around that?

I'll let Mike comment that. But I think the way I look at that, if you just take the growth part of the business and you can just -- you can play it out. Our cloud business, we've guided...

I thought you were going to say we're going to be $3 billion in cloud.

That's what I'm saying, if you just do simple math. A lot of people in this room do better math than me. So I'm an engineer by trade. I'm not a financial person. But if you just look at the trajectory of the company, of a business that is guided this year to $830 million, $840 million something of cloud ARR growing 30s and you play that out in a medium-term guide, we have a CAGR of 30-something. And you keep playing that out over the course of time with kind of no real at-scale competitor, that's a billion dollar business that is scaling very rapidly. So I look at that business in a multibillion-dollar category. And when you tie up with the overall company, not only that multibillion-dollar cloud business with a platform play with the kind of margins and cash flow we have puts us in a very unique category of [ NS1 ] in data management, in my opinion.

Mike, you've been have been with Informatica a little over 2 years. Your key conclusions and learnings about this company in the last 2 years and what are your aspirations near to midterm from a financial perspective?

It's only been 18 months, but it's been an action-packed 18 months. What stands out the most to me is the strength of the products and the uniqueness of what Informatica has to offer in what is a very dynamic and competitive marketplace. And all the good things that come from having the best products on the industry's only platform serving the multivendor, multicloud and hybrid needs of the enterprise. There's a lot of dynamism in this segment, and you're hearing about it from other companies at this conference. And because of our position, because of the strength of our products, the uniqueness of what value it provides to our customers, we're in the thick of it and we're winning in the marketplace every day, which is what gives us confidence in our growth projections, both near term and medium term for our cloud business. And furthermore, the -- our understanding of the decline part of our businesses, which are our on-prem maintenance base and our on-term license base, our understanding of that base and how it's going to decline over time and how much of it is going to move, at what pace to the cloud gives us good visibility and confidence in the ability to forecast our business. And so then to get what are my aspirations, what I see, that all blends up to a really important time right now for Informatica because this is the year when what have seen over the last 2 or 3 years because of the transition from a mixed model of on-prem and cloud to a cloud-only model, an inflection in the growth rate of ARR and revenue. So 2025 revenue and AR growth is going to be faster than '24, '26 is going to be faster than '27. And simple math would say '27 is going to be faster than '26. And over that time frame, we're going to look like a sustainable teens grower with operating margins in the high 30s to the 40s, which is going to look like any number of companies. Cadence, Synopsis, [indiscernible], PTC, Manhattan, that get very healthy multiples, much higher than what we have today. Only executing against what, to us, feels like a very visible and credible opportunity.

It's a good point. It's a good point that once we hit that magical combination of sustainable, durable growth rate, you don't try to knock it out of the ballpark. It's -- what's -- don't take risks to be a hero. Just be predictable, good growth, good margins, means higher multiple. Definitely, the certain equivalent of that outcome is not lost on it. So before I turn it over to Matt, Matt, you went to the Informatica User Conference this time, right? So Matt this guy, it was hard to get excited. Matt, how was the [indiscernible].

It was good, Kash. Yes, this AI thing and the Cloud thing is actually real. It's happening.

So on that note, I'll turn it over to you.

Let's stick with that thread then, and I don't want to butcher this, but you have a two-pronged approach to Gen AI. Informatica for Gen AI and Gen AI from Informatica. So maybe talk a little bit about both of those opportunities, if you could.

I'm glad to hear that our messaging worked for the [indiscernible]. That's my testers. I think it's a very important point because Gen AI -- when people ask about Gen AI tailwinds. Everybody starts thinking of, oh, my gosh, the GPT version of every product is the only way to monetize. That's not true for us. But Matt you just said like, look, for a Gen AI workload to be of any value, and like I said, by the way, it's it all falls apart of the data layer right now. You need to bring data from many places. You need to make sure it's cleansed, you need be at such higher quality. You need to make sure the new productionize -- it's governed, there's privacy in that. Guess what? Those are all the different capabilities that sit on our plan from IDFC. So to get your Gen AI workload to happen, you need to use those, and that's Informatica for Gen AI. Today, in that context, we were the Switzerland of data. We are now the Switzerland of AI also. We support any vector database. We support any LLM, in fact, added many, many new transformations that are needed to do a Gen AI workload, so on and so forth. So that's -- and by the way, customers are doing that today. The demo that we gave in Informatica World was a live customer use case of insurance claims processing for using Informatica for Gen AI. But we're not going to sit idle. We also have Gen AI from us to get our customers more productive, which is our Co-pilot, and our GPT. Our AI is called CLAIRE, we actually [ GA'd ] CLAIRE in 2018 when I ran products at that point. And we used it in the pre-Gen AI. It's embedded in every product. Today, every product of ours has CLAIRE Co-Pilot embedded in it, so it will give you all kinds of things that you expect from a copilot. And we also have the chat version of CLAIRE, which is called CLAIRE GPT. Chat interface, you can ask questions of CLAIRE. And because the one thing that is very secret sauce of our platform is we talked about fragmentation within an enterprise. There is no system of record for data. Will never be in my lifetime. I can guarantee you that I've seen enough tech trends. However, Google indexed the worldwide web for you to do a great search. We index the enterprise data through Meta Data. We are a system of record for Meta Data. We have almost 50 petabytes of Meta-data in the cloud. And CLAIRE GPT runs to that. When you do a search, it knows exactly what you want to find. That's the best place to go. So that's Gen AI from Informatica, and we see both of them driving Gen AI use cases and the tailwind for us to come.

So Amit, how is that adoption and usage trending? Because I know coming out of the 2Q call, you guys are giving this away for free effectively through the remainder of calendar '24. So I mean, what does that adoption rate look like? How is the utilization been across data practice?

It's pretty good. We gave CLAIRE GPT to what Matt said available to our customers in North America, go have added to the end of the year. It's a query based pricing model, which is under our consumption base pricing. Solid adoption. I think I was just talking to my IT Investor Day. About 250-plus or 275-plus customers, and we have enterprise customers. We don't look at just the users are using, querying it. And I also look at the client -- and basically, massive number of queries going through the system. We are seeing the kind of queries that's data exploration, data discovery. As an example, SSM Health, which is a provider of health care, their medical information officers who are not [indiscernible], these are business people, who are looking at making sure that I have the right professionals at the right place in the chain. So they can query and they say, here is an orthopedic -- certain set of orthopedic surgeons or orthopedic professionals, I can move them to this location today because I have a bigger demand of orthopedic patients coming over there. For pure CSAT and customer management, they're doing that real time or things like that. So we're seeing customers use that for those use cases, pretty excited. We want that to drive consumption of the platform. So for us is, give it away for this year, let adoption happen. We know the more our customers use our IPOs, the better we are in the long term.

And if I can add to that, if I may, even when we do start really charging for it, it's not going to be per se material in our financial model. So don't expect us to have a statistic next year x percent of our bookings are CLAIRE GPT or Gen AI from Informatica. The Gen AI from Informatica productivity tool, caused us to win more, enables our customers to get value out of the entire platform quicker, build more pipelines, move more data, govern more data, do more quality control, which is what really consumes the IPs. So it's tremendously valued. It's unique, but it's not the financial driver from AI. The financial driver is the Informatica for Gen AI. We talked about in our second quarter earnings call, a number of our customers that are running Gen AI pilots right now with Informatica as the data management layer, orchestrating both the app integration and the data integration to stand those use cases up. When those go into production, those are going to be big bookings. And those and many other companies doing it, that's what's going to be the AI driver that's really going to help the financial model is when those Gen AI use cases at our customers go into production.

So what are those use cases? What kind of applications are your customers building using Informatica and Gen AI?

Yes. I mean, like, for example, [ Omar's ], it's an insurance company. And basically, they have a customer sentiment -- they actually have a different one. They have a claim processing model, where the goal is that, and we showed that even at Informatica World. But today, the claims process takes months for it to settle before you know. And in the beginning, you can't even tell a customer what is the directional amount that you can actually expect coming from it. So they've created a model. We're helping them in that whole process to basically bring the claims processing down to less than a week. There's another hotel provider. Basically, they're doing customer sentiment analysis. So they basically are training their model and customer sentiment analysis. So they are basically using IDMC to bring a lot of data from within the enterprise and outside the enterprise, and put it in the models, train the model. And then we help take it out of the model and put it into what I call operational systems, whether it goes through our MBM or goes through other applications they've built. So that's -- we also had basically a supply chain use case where a technology company that does hardware production basically want to manage semiconductor company. You can just imagine that. They're part of the AI value chain, but managing the supply chain of the semi chips in that is a very, very precious commodity. So they are using our basically capability to manage the supply chain. So we're seeing very unique different front end and as well as back-office use cases for Gen AI through IDMC.

So Amit I mean as these companies train their elements on big models, is there any [ derivative ] implication within Informatica? Are you getting pulled in any of those bi-directional learning propagation?

Yes. So I think -- remember, I said something before. We are -- we will forever be the Switzerland of data. And in the world of AI, it's a data play. So we're basically the Switzerland of AI as well. So we support every LLM on every hyperscaler platform, so on and so forth. So when those deals materialize, they need for us to be there, and we are native partners with them already. Take an example of Microsoft. We are the -- when Satya stood up on stage and talked about Data Fabric, we are the native part of a Data Fabric, natively integrated design partners. There were only 2 data providers he talked about on his slide, DataBricks and Informatica. Today, Oracle is talking about it, open world, the cloud world, basically, we are the data management provider with them, and we just announced our AI blueprint with OCI. Same goes on with data by the Snowflake with the context. So we are natively involved with all the hyperscalers as well as the data platforms. And our goal is, just like in the old world, data was needed and we were sitting in the middle. We will sit in the middle and support anybody and everybody also because our customers are going to be multicloud, multi LLM, multi-vector database. So that's how we work.

So the reason I ask was you have this IP model. does it benefit as there's more learning of the data sets on part of the [ element ]?

Absolutely. I think to Mike's point, the way it goes is that more consumption, more IPOs. So today, when customers are doing experiments, they're using a current IPOs or they may buy new IPOs. But tomorrow, as those experiments start, we all talked about when does AI go into production scale. We are seeing customers doing the experiments. As they get into production, more and more IPOs get used as you get into production, and production happen in phases. It may happen in a business unit first or a geography first. So as they get into production, you need more IPUs and more IPUs and more IPO, and that's the scaling factor we will see. Well I'll say, you know what, I think the question asked, look, Gen AI is a unique thing. The build-out has happened at crazy scale. I think the best analogy I can give you is in the freeways of building in this country. The freeways got built. But the freeways have no value until the cars are driven or the driver's in the car. And I think the data cycle will be the cars and the drivers that we need to make those freeways come to fruition, and that's what we are seeing happen now.

And how close are we to production of these Gen AI applications with your customers?

But I mean it's going to be, look -- I think we have no better pick, I mean, I'll give an analogy, I may not be great at it, we're at the bottom of the first inning. By the way, when we GA'd CLAIRE in 2018, CLAIRE used to talk. So CLAIRE is not going to talk right now. It listens very well, by the way. We have very good listeners. I think -- look, I think next year. I think next year, 12 to 18 months, you will see because the reality is that we all get suckered into ChatGPT came out, the easy data available on the Internet, the card, put it in and give answers. When I'm talking to enterprises, a real mission-critical data, as they're going to put it in an experiment then they're going to take it to production. Production will be -- well, we'll productionize it in this little area. I want to make sure it doesn't go awry, if there's governance around it. And then I will scale it one more, one more. This is real hard work. It just -- you just flip the switch and then wait for the first lawsuit to happen. So I think the next 12 to 18 months is where you will start seeing, and it's no different than cloud. Remember, everybody puts storage networking into the cloud or apps into cloud. The data layer is the hard work. In AI, data layer become first. It's going to take 12 to 18 months, but I'm seeing that happen.

So you're bullish then on AI?

I think next year is going to be a year where you'll start seeing the experiments go into first phase. But I think we use the word very liberally. When we say production, now if I'm a Unilever, 96 countries, everything is -- no, that's not going to happen. They're going to put 3 countries first, then 10 countries. So I think the production is also going to be gradual. I think we've just taken the hype cycle and we've convinced ourselves that we flip a switch and everything happens. It doesn't happen like that. But I do believe the next 12 to 18 months, the productions are coming.

You made reference to DataBricks earlier, and maybe that's a good bridge into Apache, Iceberg and OpenTable file formats. There's been a lot of focus on this recently on how this could reshape the data layer. So how do you see adoption kind of impacting both Informatica's core business and kind of the broader data stack, if you will?

Yes. Well, look, first of all, for us, the more fragmentation of different technologies is better for us. the OpenTable does live by itself, right? Somebody has to feed that OpenTable. Feeding means bring data from many places, normalize it, cleanse it, do all that stuff, put it in OpenTable. And I'm simplifying that use case. And then take it from there and take it to the visualization layer of the BI layer, that's called data management. And by the way, when anything in the line system changes and it has to be constantly that [indiscernible] stay. So an OpenTable comes here and then many other things does this. So anything new is more data management. So that's a great thing for us. It's in its infancy right now, to be honest. I've not seen anybody go into that yet. It's like announcement doesn't mean adoption tomorrow. I think we oversimplify enterprises. So announcement takes some time, right? So we are going to participate in that. Just like, by the way, we saw when Hadoop came along, our Hadoop daily product went through the roof. Hadoop could move to the cloud. We move to the cloud very well. We move to the AI very well. So we participate with the more complexity we support it on. So we support OpenTable's, and we'll see good [ bets ].

Fantastic. Mike, maybe to loop you back in. You guys put out a midterm cloud ARR target of 32%, I think, fiscal '26. And I think if we reflect back on the Analyst Day, I mean, what were the assumptions you were kind of making around net new cloud business versus kind of maintenance and self managed conversion opportunity? And relative to kind of where we are today, like what's your level of conviction? And what could be kind of the upside drivers against that kind of Cloud ARR target as we kind of progress in the next couple of years?

So the build, as we think about it, for our cloud ARR and revenue growth has 2 components. One is the net new customers and workloads. We're out of the market. We're direct selling to enterprises, and we're winning against the competition. And then there's the migration growth, which is our existing installed base of maintenance customers and self-managed on-prem term license customers that move to our cloud and tend to do so at a healthy uplift in terms of what they're paying to what they're paying on the cloud. So there's 2 components. The net new component, our assumption is that we can grow at least as fast as the market is growing. And the best data we can find on how fast the cloud data management marketplace is growing is in the mid-20s. And we've shown in our investor materials the IDC buildup that gets you there. There's other ways you can triangulate around to mid-20s for market growth for cloud data management being in that level comfortably. And then the percentage points on top of that come from the migration from our existing base. We talked about how that portion has been growing even faster than the net new portion. It was about 18% of our NAR over the 12 months in December of last year. It's now about 26% of our NAR over the last 12 months ending second quarter. That may get to as high as 30% or 33%. But it's not going to be more than that. The majority of our business is going to be growing net new workloads, net new customers at or better than that market growth rate and then migrating a reasonable amount of our base to top that up into the low 30s and ultimately, by the time we get to '26, '27, the high-20.

You guys also introduced PowerCenter Cloud Edition, which I think has really helped with the migration path. So can you just talk about how that may have been accelerating the conversion opportunity across the 2 buckets of spend that you're focused on?

Yes. I mean, we introduced PowerCenter for migration. PowerCenter Cloud Edition was the round 2 of our product innovation to migrate our customers to the cloud and it came late fall last year, so it's -- this is the year it scales up. And the idea there is that like, look, our customers who are -- obviously, we've automated a lot and there was a lot more automation to be done so that this migration has lower cost, faster migration, less risk. And also, there was another one is that we wanted to give our customers the instant ability to start using the cloud. So PowerCenter Cloud Edition allows them to literally tomorrow flip the switch and start using IDMC payment [ labs ] to start operating their workloads. And then at their own pace start migrating the back end. So all of those are massive benefits or instant value creation as well as lower cost and less risk. So that came out last fall. And we obviously -- the wave model this year, we expect modernization to grow a lot faster, which we're seeing. And that, over the course of this year and next year, I expect that to drive our modernization without a doubt. We haven't talked about another thing which we also did, which is also playing out very well is that that's the tech part of it. But remember, a lot of custom work gets done around our product and everything. In the early days, we used to have us drive the services, created the blueprint. We've given it to partners. Today, almost 85%, 90% of modernization implementation work gets done by partners, which is a tremendous asset to us because they don't only do our modernization. But remember, when we get -- we modernize that stack, there's a full end-to-end modernization of the entire analytics stack that happens, so partners love that. So they have an incentive to take and learn more and drive on of that. So both of these are contributing to growth.

Questions from you guys? If you have a question, just raise your hand.

I have another one. I'm curious, like on the IDMC service specifically, right, when we think about kind of data preparation for LLMs and AI native applications, I mean, when you think about all the different components you have, whether it's data cataloging, data classification, governance, privacy, I mean, is there any certain area where you expect green shoots predicated on kind of this broader secular theme?

Maybe 3 steps. First, obviously, like it's the plumbing that needs to happen first, okay, the integration capabilities, quality capabilities, those kind of -- and there are so many in app data. So those are the first things to move around data, normalize, clean it and everything, and cataloging comes in the middle. I think enterprises in parallel start looking after the experiment start. Okay, if I have to do step one of some operationalization, I need to put governance on top. Because you need to have the guardrails otherwise, it can go nuts. So governance will come around that. And the third one is that as you start then productionizing and do some real action-oriented work with it, MDM plays into that. Because if I am doing a customer sentiment analysis, MDM is used as a customer 360. That may be the last mile to do action. So I think it's that phasing in which we see the workloads start happening.

I have a question. I know you talked about being part of the central part of the data stack, but you get people like Fivetran and DBT and all the others who will want to be part of the modern data stack. So I'd like to understand as an investor how they all fit together?

Sure. By the way, so I think the word -- we have to separate those players from modern data stack. We are the modern data stack. I mean, the reality is that my cloud serves the modern data stack. Most of the work -- almost all the workload is Snowflake related, data base-related, Azure-, AWS-related, right? So that's the one thing I want to clear that our cloud is not serving some old data stack. We are serving the modern data stack. Our cloud business is not tied to the mainframe. So I think let's just kind of throw that out of the way. So now let's come to the modern data stack. The modern data stack requires many services. Last I checked for -- and by the way, if you would really, really want to see that, just click on YouTube go to My Informatica World, you'd see the description of the modern data stack. To stitch that modern data stack, you need easily, easily 30-plus data management providers, Fivetran, DBT, Reltio, Elation, Calibra and on and on and on. And sure, if you want to do a point project, God Bless you, that's fine and it's like, okay, if you went to do something cheap. But if I'm a large enterprise and I'm building a modern data stack, if I have to stitch 30, not only it's more cost. If 2 of them go bankrupt, 1 of them has a different road map than the other one. You will never have a data, and we see that a lot. So we have all of those services. And lastly, check the Magic Quadrant. We are leader against all of them. So modern doesn't mean that you are a start-up. Modern means that you build the best product.

I've been asking every 1 of the 12 or so CEOs that we had a chance to interview in the last day, 1.5 days, and that is what are you doing with Gen AI internally? Have you seen any benefits from internal deployments of Gen AI within Informatica?

Yes, early innings we are. So clearly, I think -- everybody says early innings. I mean, because I think everybody thinks that -- I think we've gotten -- and I'll take a side. People have gotten super confused about this semiconductor buildout and as if that's the piece, everything happens. Look, guys, I mean, like one provider has the GPUs. There are 3 hyperscalers who have the hundreds of billion dollars of cash. They don't know what to do with it. So by the way, if the data centers are not in the cloud cash, Goldman Sachs will not be buying GPUs that at the pace at which hyperscalers are buying. Let's be very clear, right? So I think there's a very [ anomalous ] build out, build it and it will come. We are using it. My marketing team is using it. Our customer success team is using it. We have our own implementations of Gen AI. R&D in its early stages. And remember, infrastructure software is a bit of a tricky piece of software. Industrial-grade software users where you print your 10-K, 10-Q reports, you give stress test data to the ECB as an example. If I have a GitHub copilot generated code and it gives you $1 million different than what the ECB wants to see, you're going to get sued. So we've got to be careful, right? So this is not -- so I think we have this fascination to take something which is instantly -- so we are absolutely experimenting. But we have areas where it is very low cost. By the way, we've also started using CLAIRE GPT for a very different use case. Our customers are using CLAIRE GPT, which we use that to connect to our internal knowledge docs to search up how to use Informatica products. That is an efficiencies of they're not calling customer support. So we are seeing those implementations. Our goal is to create more of those. But again, I say less because -- small because it's not game-changing tomorrow. But I think as you're talking to CEOs, I talked a bunch of CEOs who are part of a CEO forum. And everybody has a desire to do what we will do more, but it takes it on pace.

Mike, will that matter to the bottom line? You've got one second. You had more than one second.

It will. How material it will be when you notice it probably won't be until '26, to be fair. But in '25, I think we're going to find some of those customer success, marketing, even go-to-market. We're experimenting with using Gen AI DDRs, right? Some of those are going to come to fruition and will enable us to get more operating leverage out of the model as we grow in '26 and beyond.

And any providers that come to mind on which technologies your folks are internally leveraging?

We're expending everybody.

Everybody?

Yes. Because everybody has -- different people have strength in different areas. Not everybody -- there's no one multipurpose best in every...

So Windows, Copilot, Office Copilot, nothing -- one thing is...

Not just. There's no silver bullet.

No Silver bullet. On that note, thank you once again. It's always a very energetic discussion. I really enjoy hanging on.

I hope I [indiscernible].

I think you did. Yes.

Thank you. Thanks for hosting.

Thanks for coming in.